Mercurial > hg
view mercurial/worker.py @ 42044:bb271ec2fbfb
compression: introduce a `storage.revlog.zstd.level` configuration
This option control the zstd compression level used when compressing revlog
chunk. The usage of zstd for revlog compression has not graduated from
experimental yet, but we intend to fix that soon.
The option name for the compression level is more straight forward to pick, so
this changesets comes first. Having a dedicated option for each compression
engine is useful because they don't support the same range of values.
I ran the same measurement as for the zlib compression level (in the parent
changesets). The variation in repository size is stay mostly in the same (small)
range. The "read/write" performance see smallish variation, but are overall much
better than zlib. Write performance show the same tend of having better write
performance for when reaching high-end compression.
Again, we don't intend to change the default zstd compression level (currently:
3) in this series. However this is worth investigating in the future.
The Performance comparison of zlib vs zstd is quite impressive. The repository
size stay in the same range, but the performance are much better in all
situations.
Comparison summary
==================
We are looking at:
- performance range for zlib
- performance range for zstd
- comparison of default zstd (level-3) to default zlib (level 6)
- comparison of the slowest zstd time to the fastest zlib time
Read performance:
-----------------
| zlib | zstd | cmp | f2s
mercurial | 0.170159 - 0.189219 | 0.144127 - 0.149624 | 80% | 88%
pypy | 2.679217 - 2.768691 | 1.532317 - 1.705044 | 60% | 63%
netbeans | 122.477027 - 141.620281 | 72.996346 - 89.731560 | 58% | 73%
mozilla | 147.867662 - 170.572118 | 91.700995 - 105.853099 | 56% | 71%
Write performance:
------------------
| zlib | zstd | cmp | f2s
mercurial | 53.250304 - 56.2936129 | 40.877025 - 45.677286 | 75% | 86%
pypy | 460.721984 - 476.589918 | 270.545409 - 301.002219 | 63% | 65%
netbeans | 520.560316 - 715.930400 | 370.356311 - 428.329652 | 55% | 82%
mozilla | 739.803002 - 987.056093 | 505.152906 - 591.930683 | 57% | 80%
Raw data
--------
repo alg lvl .hg/store size 00manifest.d read write
mercurial zlib 1 49,402,813 5,963,475 0.170159 53.250304
mercurial zlib 6 47,197,397 5,875,730 0.182820 56.264320
mercurial zlib 9 47,121,596 5,849,781 0.189219 56.293612
mercurial zstd 1 49,737,084 5,966,355 0.144127 40.877025
mercurial zstd 3 48,961,867 5,895,208 0.146376 42.268142
mercurial zstd 5 48,200,592 5,938,676 0.149624 43.162875
mercurial zstd 10 47,833,520 5,913,353 0.145185 44.012489
mercurial zstd 15 47,314,604 5,728,679 0.147686 45.677286
mercurial zstd 20 47,330,502 5,830,539 0.145789 45.025407
mercurial zstd 22 47,330,076 5,830,539 0.143996 44.690460
pypy zlib 1 370,830,572 28,462,425 2.679217 460.721984
pypy zlib 6 340,112,317 27,648,747 2.768691 467.537158
pypy zlib 9 338,360,736 27,639,003 2.763495 476.589918
pypy zstd 1 362,377,479 27,916,214 1.532317 270.545409
pypy zstd 3 354,137,693 27,905,988 1.686718 294.951509
pypy zstd 5 342,640,043 27,655,774 1.705044 301.002219
pypy zstd 10 334,224,327 27,164,493 1.567287 285.186239
pypy zstd 15 329,000,363 26,645,965 1.637729 299.561332
pypy zstd 20 324,534,039 26,199,547 1.526813 302.149827
pypy zstd 22 324,530,595 26,198,932 1.525718 307.821218
netbeans zlib 1 1,281,847,810 165,495,457 122.477027 520.560316
netbeans zlib 6 1,205,284,353 159,161,207 139.876147 715.930400
netbeans zlib 9 1,197,135,671 155,034,586 141.620281 678.297064
netbeans zstd 1 1,259,581,737 160,840,613 72.996346 370.356311
netbeans zstd 3 1,232,978,122 157,691,551 81.622317 396.733087
netbeans zstd 5 1,208,034,075 160,246,880 83.080549 364.342626
netbeans zstd 10 1,188,624,176 156,083,417 79.323935 403.594602
netbeans zstd 15 1,176,973,589 153,859,477 89.731560 428.329652
netbeans zstd 20 1,162,958,258 151,147,535 82.842667 392.335349
netbeans zstd 22 1,162,707,029 151,150,220 82.565695 402.840655
mozilla zlib 1 2,775,497,186 298,527,987 147.867662 751.263721
mozilla zlib 6 2,596,856,420 286,597,671 170.572118 987.056093
mozilla zlib 9 2,587,542,494 287,018,264 163.622338 739.803002
mozilla zstd 1 2,723,159,348 286,617,532 91.700995 570.042751
mozilla zstd 3 2,665,055,001 286,152,013 95.240155 561.412805
mozilla zstd 5 2,607,819,817 288,060,030 101.978048 505.152906
mozilla zstd 10 2,558,761,085 283,967,648 104.113481 497.771202
mozilla zstd 15 2,526,216,060 275,581,300 105.853099 591.930683
mozilla zstd 20 2,485,114,806 266,478,859 95.268795 576.515389
mozilla zstd 22 2,484,869,080 266,456,505 94.429282 572.785537
author | Pierre-Yves David <pierre-yves.david@octobus.net> |
---|---|
date | Wed, 27 Mar 2019 18:35:59 +0100 |
parents | e10adebf8176 |
children | 5ca136bbd3f6 |
line wrap: on
line source
# worker.py - master-slave parallelism support # # Copyright 2013 Facebook, Inc. # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. from __future__ import absolute_import import errno import os import signal import sys import threading import time try: import selectors selectors.BaseSelector except ImportError: from .thirdparty import selectors2 as selectors from .i18n import _ from . import ( encoding, error, pycompat, scmutil, util, ) def countcpus(): '''try to count the number of CPUs on the system''' # posix try: n = int(os.sysconf(r'SC_NPROCESSORS_ONLN')) if n > 0: return n except (AttributeError, ValueError): pass # windows try: n = int(encoding.environ['NUMBER_OF_PROCESSORS']) if n > 0: return n except (KeyError, ValueError): pass return 1 def _numworkers(ui): s = ui.config('worker', 'numcpus') if s: try: n = int(s) if n >= 1: return n except ValueError: raise error.Abort(_('number of cpus must be an integer')) return min(max(countcpus(), 4), 32) if pycompat.isposix or pycompat.iswindows: _STARTUP_COST = 0.01 # The Windows worker is thread based. If tasks are CPU bound, threads # in the presence of the GIL result in excessive context switching and # this overhead can slow down execution. _DISALLOW_THREAD_UNSAFE = pycompat.iswindows else: _STARTUP_COST = 1e30 _DISALLOW_THREAD_UNSAFE = False def worthwhile(ui, costperop, nops, threadsafe=True): '''try to determine whether the benefit of multiple processes can outweigh the cost of starting them''' if not threadsafe and _DISALLOW_THREAD_UNSAFE: return False linear = costperop * nops workers = _numworkers(ui) benefit = linear - (_STARTUP_COST * workers + linear / workers) return benefit >= 0.15 def worker(ui, costperarg, func, staticargs, args, threadsafe=True): '''run a function, possibly in parallel in multiple worker processes. returns a progress iterator costperarg - cost of a single task func - function to run staticargs - arguments to pass to every invocation of the function args - arguments to split into chunks, to pass to individual workers threadsafe - whether work items are thread safe and can be executed using a thread-based worker. Should be disabled for CPU heavy tasks that don't release the GIL. ''' enabled = ui.configbool('worker', 'enabled') if enabled and worthwhile(ui, costperarg, len(args), threadsafe=threadsafe): return _platformworker(ui, func, staticargs, args) return func(*staticargs + (args,)) def _posixworker(ui, func, staticargs, args): workers = _numworkers(ui) oldhandler = signal.getsignal(signal.SIGINT) signal.signal(signal.SIGINT, signal.SIG_IGN) pids, problem = set(), [0] def killworkers(): # unregister SIGCHLD handler as all children will be killed. This # function shouldn't be interrupted by another SIGCHLD; otherwise pids # could be updated while iterating, which would cause inconsistency. signal.signal(signal.SIGCHLD, oldchldhandler) # if one worker bails, there's no good reason to wait for the rest for p in pids: try: os.kill(p, signal.SIGTERM) except OSError as err: if err.errno != errno.ESRCH: raise def waitforworkers(blocking=True): for pid in pids.copy(): p = st = 0 while True: try: p, st = os.waitpid(pid, (0 if blocking else os.WNOHANG)) break except OSError as e: if e.errno == errno.EINTR: continue elif e.errno == errno.ECHILD: # child would already be reaped, but pids yet been # updated (maybe interrupted just after waitpid) pids.discard(pid) break else: raise if not p: # skip subsequent steps, because child process should # be still running in this case continue pids.discard(p) st = _exitstatus(st) if st and not problem[0]: problem[0] = st def sigchldhandler(signum, frame): waitforworkers(blocking=False) if problem[0]: killworkers() oldchldhandler = signal.signal(signal.SIGCHLD, sigchldhandler) ui.flush() parentpid = os.getpid() pipes = [] for pargs in partition(args, workers): # Every worker gets its own pipe to send results on, so we don't have to # implement atomic writes larger than PIPE_BUF. Each forked process has # its own pipe's descriptors in the local variables, and the parent # process has the full list of pipe descriptors (and it doesn't really # care what order they're in). rfd, wfd = os.pipe() pipes.append((rfd, wfd)) # make sure we use os._exit in all worker code paths. otherwise the # worker may do some clean-ups which could cause surprises like # deadlock. see sshpeer.cleanup for example. # override error handling *before* fork. this is necessary because # exception (signal) may arrive after fork, before "pid =" assignment # completes, and other exception handler (dispatch.py) can lead to # unexpected code path without os._exit. ret = -1 try: pid = os.fork() if pid == 0: signal.signal(signal.SIGINT, oldhandler) signal.signal(signal.SIGCHLD, oldchldhandler) def workerfunc(): for r, w in pipes[:-1]: os.close(r) os.close(w) os.close(rfd) for result in func(*(staticargs + (pargs,))): os.write(wfd, util.pickle.dumps(result)) return 0 ret = scmutil.callcatch(ui, workerfunc) except: # parent re-raises, child never returns if os.getpid() == parentpid: raise exctype = sys.exc_info()[0] force = not issubclass(exctype, KeyboardInterrupt) ui.traceback(force=force) finally: if os.getpid() != parentpid: try: ui.flush() except: # never returns, no re-raises pass finally: os._exit(ret & 255) pids.add(pid) selector = selectors.DefaultSelector() for rfd, wfd in pipes: os.close(wfd) selector.register(os.fdopen(rfd, r'rb', 0), selectors.EVENT_READ) def cleanup(): signal.signal(signal.SIGINT, oldhandler) waitforworkers() signal.signal(signal.SIGCHLD, oldchldhandler) selector.close() return problem[0] try: openpipes = len(pipes) while openpipes > 0: for key, events in selector.select(): try: yield util.pickle.load(key.fileobj) except EOFError: selector.unregister(key.fileobj) key.fileobj.close() openpipes -= 1 except IOError as e: if e.errno == errno.EINTR: continue raise except: # re-raises killworkers() cleanup() raise status = cleanup() if status: if status < 0: os.kill(os.getpid(), -status) sys.exit(status) def _posixexitstatus(code): '''convert a posix exit status into the same form returned by os.spawnv returns None if the process was stopped instead of exiting''' if os.WIFEXITED(code): return os.WEXITSTATUS(code) elif os.WIFSIGNALED(code): return -os.WTERMSIG(code) def _windowsworker(ui, func, staticargs, args): class Worker(threading.Thread): def __init__(self, taskqueue, resultqueue, func, staticargs, *args, **kwargs): threading.Thread.__init__(self, *args, **kwargs) self._taskqueue = taskqueue self._resultqueue = resultqueue self._func = func self._staticargs = staticargs self._interrupted = False self.daemon = True self.exception = None def interrupt(self): self._interrupted = True def run(self): try: while not self._taskqueue.empty(): try: args = self._taskqueue.get_nowait() for res in self._func(*self._staticargs + (args,)): self._resultqueue.put(res) # threading doesn't provide a native way to # interrupt execution. handle it manually at every # iteration. if self._interrupted: return except pycompat.queue.Empty: break except Exception as e: # store the exception such that the main thread can resurface # it as if the func was running without workers. self.exception = e raise threads = [] def trykillworkers(): # Allow up to 1 second to clean worker threads nicely cleanupend = time.time() + 1 for t in threads: t.interrupt() for t in threads: remainingtime = cleanupend - time.time() t.join(remainingtime) if t.is_alive(): # pass over the workers joining failure. it is more # important to surface the inital exception than the # fact that one of workers may be processing a large # task and does not get to handle the interruption. ui.warn(_("failed to kill worker threads while " "handling an exception\n")) return workers = _numworkers(ui) resultqueue = pycompat.queue.Queue() taskqueue = pycompat.queue.Queue() # partition work to more pieces than workers to minimize the chance # of uneven distribution of large tasks between the workers for pargs in partition(args, workers * 20): taskqueue.put(pargs) for _i in range(workers): t = Worker(taskqueue, resultqueue, func, staticargs) threads.append(t) t.start() try: while len(threads) > 0: while not resultqueue.empty(): yield resultqueue.get() threads[0].join(0.05) finishedthreads = [_t for _t in threads if not _t.is_alive()] for t in finishedthreads: if t.exception is not None: raise t.exception threads.remove(t) except (Exception, KeyboardInterrupt): # re-raises trykillworkers() raise while not resultqueue.empty(): yield resultqueue.get() if pycompat.iswindows: _platformworker = _windowsworker else: _platformworker = _posixworker _exitstatus = _posixexitstatus def partition(lst, nslices): '''partition a list into N slices of roughly equal size The current strategy takes every Nth element from the input. If we ever write workers that need to preserve grouping in input we should consider allowing callers to specify a partition strategy. mpm is not a fan of this partitioning strategy when files are involved. In his words: Single-threaded Mercurial makes a point of creating and visiting files in a fixed order (alphabetical). When creating files in order, a typical filesystem is likely to allocate them on nearby regions on disk. Thus, when revisiting in the same order, locality is maximized and various forms of OS and disk-level caching and read-ahead get a chance to work. This effect can be quite significant on spinning disks. I discovered it circa Mercurial v0.4 when revlogs were named by hashes of filenames. Tarring a repo and copying it to another disk effectively randomized the revlog ordering on disk by sorting the revlogs by hash and suddenly performance of my kernel checkout benchmark dropped by ~10x because the "working set" of sectors visited no longer fit in the drive's cache and the workload switched from streaming to random I/O. What we should really be doing is have workers read filenames from a ordered queue. This preserves locality and also keeps any worker from getting more than one file out of balance. ''' for i in range(nslices): yield lst[i::nslices]